Example 1: an.
Example 2: an.
Example 3: anx.
Example 4: xi + xi+1.
Example 5: x2.
The table below provides details on how different screen readers and talking browsers output the formulas above.
| Screen reader and/or browser | 1: subscript | 2: superscript | 3: superscript and subscript | 4: subscript with sum | 5: x squared |
|---|---|---|---|---|---|
| CLC-4-TTS Suite with Fire Vox - Bundle Pack 3.3: Auto Read | example 1 (pause) a n | example 2 (pause) a n | example 3 (pause) a n x | example 4 (pause) x i plus x i plus one | example 5 (pause) x two |
| CLC-4-TTS Suite with Fire Vox - Bundle Pack 3.3: Read Selected | example 1 (pause) an | example 2 (pause) an | example 3 (pause) anks | example 4 (pause) zi plus zi plus one | example 5 (pause) x two |
| Internet Explorer 7 and JAWS 8 on Parallels (Windows Media video) | example 1 colon-an | example 2 colon-an | example 3 colon-anks | example 4 colon zi plus zi plus one | example 5 colon x two |
| Internet Explorer 6 with JAWS 8 | JAWS 8 fails to read the examples. If one reads the line character by character, it does announce the markup only if a digit is in the sub / sup and that too only for the digit. In example 4 for instance, where x+1 is marked up, it only announces the digit 1 as being marked up. It reads example 5 ok in character mode. | ||||
| Internet Explorer 7 with JAWS 7.10 and JAWS 8.0 (speech) | Both JAWS 7.10 and JAWS 8 fail to read all examples that have alphabets within sub or sup.
Only the digits are announced with their corresponding sub / sup markup when one reads the text in character mode navigating with right/left arrow keys.
So this is the same for JAWS 8 with IE 6 (reported on Sept 14) but not for JAWS 7.1 and IE 6. |
||||
| JAWS 7.10 with Internet Explorer 7 (speech: MP3) | example 1 an | example 2 an | example 3 anks (or x?) | example 4 zi plus zi plus one | example 5 x two |
| JAWS 7.10 with Internet Explorer 7: Braille | Example: 1 an | Example: 2 an | Example 3: anx | Example 4: xi + xi+1 | Example 5: x2 |
| Opera 9.5 alpha build 14 September 2007 with VoiceOver on MacOS X 10.4.10 | It just reads the superscripts and subscripts as seperated blocks with no particular disctinction or announcement of how they differ. | ||||
| Safari 2.04 with VoiceOver on MacOS X 10.4.10 | It just reads the superscripts and subscripts as seperated blocks with no particular disctinction or announcement of how they differ. | ||||
| Internet Explorer 7 with Windows-Eyes 5.5 | It just reads the superscripts and subscripts as seperated blocks with no particular disctinction or announcement of how they differ. There seem to be no settings in Windows-Eyes that make it announce the super/subscript differently. | ||||
| Firefox 3 beta (16 September 2007) and Orca (from Subversion repository, 15 September 2007) on Linux Gnome | There is no information on the superscript or subscript. | ||||
For background, see the thread on the WAI Interest Group mailing list and the thread on the WCAG Working Group mailing list.
When the test page is saved as text, different browsers generate different representations of the formulas.
| Browser | 1: subscript | 2: superscript | 3: superscript and subscript | 4: subscript with sum | 5: x squared |
|---|---|---|---|---|---|
| Firefox 2.0, SeaMonkey 1.1 | Example 1: a_n . | Example 2: a^n . | Example 3: a_n ^x . | Example 4: x_i + x_i+1 . | Example 5: x^2 . |
| Internet Explorer 6 | Example 1: an. | Example 2: an. | Example 3: anx. | Example 4: xi + xi+1. | Example 5: x2. |
| Opera 9.02 | Example 1: an. | Example 2: an. | Example 3: anx. | Example 4: xi + xi+1. | Example 5: x2. |
| Lynx version 2.8.4 rel 1 (2002) on Debian Sarge | Example 1: a[n]. | Example 2: a^n | Example 3: a[n]x. | Example 4: x[i] + x[i+1]. | Example 5: x2. |
Many thanks to Joshue O’Connor, Sailesh Panchang, Charles McCathieNevile, Steve Faulkner, Osvaldo La Rosa, Jonathan Chacón, Alan Chuter and Bart Simons for sending test results.
TinyURL: http://tinyurl.com/2qll7h.